Differential impact of monetary policy: a sectoral approach

The proposition tested in the thesis is that actual money growth is not systematically related to output growth of selected manufacturing sectors. Given that money growth is found to be non-neutral, two further tests are undertaken.

First, I test the rational expectations hypothesis that only unexpected monetary changes affect output growth against the Phillips' type stabilization hypothesis that both expected and unexpected monetary growth affect output growth. Further, I examine the arguments that under different conditions the degree of association between output growth and money growth (given statistically by the magnitude of the coefficient on money growth) is not constant. The conditions tested relate to the level of capacity utilization and the phase of the business cycle.

Secondly, I test for differential sector to aggregate manufacturing output response. Non-neutrality does not imply differentiality since money growth can be significantly related to individual manufacturing sectors in an identical way. Differential effects might be argued to exist because the different structure of each sector imposes different restrictions on decisions and actions that can be undertaken. If differential effects are found to exist, a test of specific sector characteristics, which may be functionally related to the differential association, will be undertaken.

Money growth was found to be significantly related to individual sector output growth in equations unadjusted for seasonal variation. No association was found in seasonally adjusted estimations indicating that money growth is related to the seasonal amplitude and not the underlying non-seasonal structure. This supports the neutrality proposition.

Expected and unexpected money growth, in the seasonally unadjusted estimations, were found to be statistically significant in nine of the twelve industries examined. Further, these associations are strongly related to the level of capacity utilization and the phase of the business cycle.

The test of differential impact did not support the hypothesis that the incidence of money growth was unequal across sectors. Therefore, no further investigation was necessary.